EBU Technical Review

No. 300 (October 2004)

Convergence?

The prospect of “convergence” was a hot topic in the mid-1990s – with many people confidently predicting the rise of mega-industries that would unite separate segments of industry, such as publishing, broadcasting, telecoms and information technology. In practice, this form of convergence did not really happen – apart from the infamous merger of AOL and Time Warner in January 2000, in which the upstart AOL effectively took over the more traditional publishing and media empire. At that time, many analysts predicted that this was just the first of many such mergers. In retrospect, the merger was not successful – to the extent that, in September 2003, AOL Time Warner's board of directors decided to remove the “AOL” from its corporate name.

The idea of convergence of industries may have been killed off by the bursting of the Internet bubble, but what about other aspects of convergence, such as convergence of consumer terminals? I have always doubted that people will use their TV sets for word processing, but conversely I have long believed that computers will be used as TV receivers. Much attention at IBC 2004 in Amsterdam was focussed on handheld devices displaying TV-like services but, for me, one of the best exhibits at IBC 2004 involved DVB-T reception on a lap-top computer: reception of the Dutch DVB-T services was “rock solid” even when I waved the lap-top above my head during one of the sessions. I must admit that this prompted me to say “Eat your hearts out, ATSC!” because we know that our friends in ATSC simply cannot match that level of performance. Of course, computers can provide added functionality beyond TV reception, such as the ability to record TV programmes on the computer’s hard disk.

In the digital world, a “bit” is a “bit” whether it represents audio, video, data, software, etc. Consequently, in principle, there is no reason for separate networks for broadcasting and telecommunications. Of course, the reality is slightly more complex than many non-technical people might believe! In practice, there are some fundamental distinctions between broadcasting and telecommunications. Broadcasting essentially delivers one-way, one-to-many services, whereas telecoms operators provide two-way, one-to-one services. Nevertheless, the boundaries between broadcasting and telecoms have become increasingly blurred in recent years: since the early 1980s, some analogue TV services have been used to deliver limited “telecoms-like” services, such as delivery of encrypted teletext services for individuals or closed groups of users. The introduction of digital radio and TV networks opens up new opportunities for data services, particularly for delivery to portable or hand-held devices. Furthermore, telecoms operators now expect that consumers will pay, say, €1.50 to watch a 30-second video clip (e.g. a football goal) on their mobile phones. Despite the fact that some telecoms operators paid huge amounts of money for access to UMTS spectrum, current wisdom suggests that it will not be economic to deliver large-scale on-demand video services via UMTS. Instead, attention has turned to the benefits of one-to-many services: rather than sending individual video streams to each consumer, it would be much more efficient to transmit the same material simultaneously to all those interested. Of course, this “new idea” is actually “broadcasting”!

In the area of “convergent” delivery systems, the most important developments are DVB-H and DMB. DVB-H is an extension of the DVB-T system designed for reception on handheld devices, whereas DMB is an extension of the DAB system. Which system is best? To date, there is no definitive answer to that question, but I suggest that you should read very carefully the article by Andreas Sieber and Chris Weck in the July 2004 issue of EBU Technical Review.

Broadcasters should welcome the enthusiasm of telecoms operators to use broadcast technologies, such as DVB-H or DMB. However, they should also recognise that the telecoms operators hope to use frequency bands currently used for “traditional” broadcasting. From my perspective, frequency availability will be the biggest obstacle to the success of DVB-H. It will be difficult in most countries to find UHF TV channels that can be used on a widespread basis for DVB-H services. Those readers with long memories will recall that the Eureka-147 DAB system was originally devised to use an entire TV channel (i.e. a bandwidth of 7 or 8 MHz), but it was eventually decided that the DAB system should use channels with 1.5 MHz bandwidth. This very wise decision meant that one 7 MHz channel could support 4 DAB blocks, thus giving much more flexibility to the frequency planners.

Another relevant subject was considered in an article by Simon Mason in the July 2004 issue of the Technical Review. This article gave details of building penetration loss for DAB services in VHF Band III, which showed that reliable reception in buildings requires much higher field strengths than normally considered for DAB planning. In practice, the increased field strength would be best achieved by using a dense network of low-power transmitters – more like a cellular network than the traditional large service areas provided by relatively high-power broadcast transmitters. As building penetration loss increases with frequency, we can expect this problem to be even more serious in UHF Bands IV and V, thus exacerbating the frequency problems for DVB-H.

Digital broadcasting uses the radio spectrum much more efficiently than analogue broadcasting, thus leading to the idea of a “digital dividend” – which implies that when the conversion process from analogue to digital broadcasting has been completed, some spectrum previously used for broadcasting can be released for other applications.

Some countries are hoping for a large digital dividend: for example, regulators in the UK have announced that they plan to release 14 UHF TV channels (a total of 112 MHz) when analogue TV services are closed down. Other countries are more cautious, usually because they cannot find enough spectrum to offer widespread coverage of digital terrestrial TV. The experience to date suggests that the success of digital terrestrial TV requires a wide range of new digital services, together with digital simulcasts of existing analogue terrestrial services. If hardly any new services are offered, consumers have little incentive to transfer from analogue to digital. Obviously, if more digital TV services are offered, the potential digital dividend will be reduced.

Another issue is whether digital terrestrial TV should be based on set-top antennas or, even, mobile receivers. Mobile and portable reception is clearly attractive – but also requires significantly more spectrum, again reducing the digital dividend. A further factor is the role of HDTV: this year has seen a resurgence of interest in HDTV in Europe. As the ITU’s 1961 Stockholm Plan has had a lifetime of more than 40 years, the replacement Plan (see article in this issue) to be agreed in May 2006 (at the ITU Regional Radiocommunication Conference RRC-06) will probably need to be valid for, at least, 30 years. Even those who are currently sceptical about the case for HDTV accept that HDTV will eventually become the norm – in the same way that colour TV has become almost universal (it is now very difficult to purchase black-and-white TVs). How much spectrum should be reserved today to permit the introduction of HDTV broadcast services?

The full benefits of digital terrestrial TV in terms of spectrum efficiency are dependent on the complete withdrawal of analogue broadcasting, not just within a given country but also in adjacent countries. Given the varied timetables for the digital transition within European countries, the digital dividend will become available much later than many currently hope. After RRC-06, we will have a much better idea of the size of the digital dividend. As all countries in Europe (with the exception of Iceland) will be constrained by their neighbours’ frequency plans, it would be prudent to wait for RRC-06 before making irrevocable decisions about how to spend the digital dividend. In other words, wait for the lottery result before spending your winnings!

* * * * * * *

This is the 300th issue of EBU Technical Review. Dramatic changes in technology have taken place since the EBU was established in February 1950. In the first issue, published in May 1950, the then president of the EBU, Ian Jacob, wrote:

"The coming into force of the Copenhagen Plan on March 15th provided very early in its career a test for the Technical Centre, and gave it the chance to demonstrate its effectiveness. Everything points to the conclusion that the European Broadcasting Union has a useful future before it, …".

For our younger readers, I should explain that the 1948 Copenhagen Plan was concerned with frequency assignments in the LF and MF bands. Despite the passage of time, frequency-planning conferences continue to be of great importance to EBU Members – as indicated by the above Editorial, together with the article in the current issue on ITU RRC-04. This prompts me to remark “The more things change, the more they stay the same.”

Since the early days, the objective of the Technical Review has been to publish articles that are comprehensive and authoritative – whilst trying to ensure that such articles are accessible to non-specialists. Indeed, in 1958, it was specified that:

“We want to obviate the danger of having articles which mean nothing except to their authors and a few specialists who might have written them themselves!”

It is not for me to judge how well we have achieved these objectives, but feedback from our readers continues to be very positive. As the success of the Technical Review is entirely dependent on our authors, I would like to take this opportunity to thank all of those who have contributed articles in the past and to encourage them and others to provide us with high-quality content in the future.

Philip Laven
Director
EBU Technical Department

Ed. The Technical Review started off as the EBU Document and Information Bulletin in May 1950. In January 1958 (edition 47), it became the E.B.U. Review – Technical. It was given its current name in Spring 1992 (edition 251). The Contents pages and Editorials from these landmark editions are reproduced as scanned image files in a PDF file (742 KB) that you can find here.

Editorials Archive